This diagnosis is seductively simple–there is too much capital in the industry. In addition, the therapeutic regimen implicit in the “excess capital diagnosis” avoids some of the most bitter medicine of all–that is, the need for carriers and agents to make fundamental changes to their workers' comp operational strategies.

Not convinced that excessive competition isnt the root cause of poor loss experience? Take a look at Chart 1, showing seven selected states ratio of written to standard premium–a commonly used barometer of pricing posture. This pricing index varies from a low of .68 in Texas to a high of 1.08 in Florida.

The experience in Texas and Illinois is consistent with the soft pricing theory (the weakest pricing resulting in the highest loss ratio), as is the inverse, in the experience in Wisconsin and Florida. However, the significant variation in loss ratios still exist–for example, why was Indianas pricing so low and loss experience so good? These discrepancies suggest that other factors must be at work.

You may well ask, whats the missing factor? Lets look at a couple of possibilities.

Looking at Chart 2, for example, we can see just how much these states were out of sync with each other in terms of employment growth–and just as importantly, the composition of this growth–over the last decade. In some states, growth accelerated late in the decade; in others, it decelerated. While Texas and Colorado sustained very strong growth throughout the decade, Illinois and Michigan experienced relatively muted but stable growth.

Strong growth in construction-related employment is one of the most important aspects of a states economic growth–in part due to the high exposures and premiums generated by construction-related classes.

As shown in Chart 3, construction employment grew very rapidly in many but not all of the selected states in the second half of the 1990s. In particular, Colorado saw exceptionally fast growth in construction–enough to raise constructions share of total employment by almost a full two percentage points.

Colorados surge in construction activity produced a sharp jump in the number of small, new or relatively new construction businesses needing workers' comp coverage. Most of these accounts are either too new or too small to be experience-rated, and as a result are written using filed rates reflecting the aggregate loss experience of the rating class.

Why does this matter? In contrast with the class-average approach used in rate making, Occupational Safety and Health Administration data indicate that smaller construction accounts tend to have significantly higher accident frequency than their larger counterparts.

As a result, a surge in construction activity tends to introduce a substantial number of under-rated smaller risks into the states workers' comp system. This exacerbates the states loss experience without any change in carriers pricing posture.

A surge in construction activity also boosts employers reliance on overtime to meet delivery timetables–at least in the short-run until employment levels can equilibrate. Accident frequency is likely to rise as well due to worker fatigue and lessened focus on safe work practices.

Overtime pay differentials do boost premium collections, helping to compensate for increased accident frequency. However, it is likely that substantial overtime hours impact on loss potential extends throughout the workweek, and therefore, is not fully reflected in the premium generated from overtime pay differentials.

Lets look at another possible factor to explain loss ratios–variations in employees access to health insurance coverage. As shown in Chart 4, health insurance coverage varies significantly between classes of business and between states.

In both Florida and the entire United States, the construction industry has a very low rate of insurance coverage, second only to the agriculture industry. As a states construction employment surges, an increased fraction of uninsured new employees is introduced into the states workers' comp system, with an adverse impact on the frequency of questionable workers' comp claims.

Heres a third possible factor. The ups and downs of each states economy also influence job experience as workers take advantage of new positions–or, in retractions, experience layoffs.

In addition, some industries historically have experienced much higher job turnover than others. For agriculture and retail trade, the high percentage of job entrants in part reflects the greater prevalence of unskilled positions.

Within the United States as a whole, the construction industry also attracts a relatively high level of inexperienced workers. As a result, construction booms tend to significantly increase the share of inexperienced employees, thereby deteriorating loss experience in the state.

Reliance on more youthful, less experienced workers during periods of strong economic growth also can exacerbate loss experience. Agriculture and retail trade tend to make heaviest use of young employees. However, cyclically sensitive industries such as construction tend to make increased use of more accident-prone, younger employees during periods of fast growth such as Texas and Colorado experienced in the mid-to-late-1990s.

The entrance of these young workers tends to drive up the actual accident frequency of certain classes relative to the frequency implicit in the loss experience used for rate making.

Youthful workers, heavy overtime, small firms, lack of health insurance, and high job turnover–these factors constitute a diagnostic list that would make even H.L. Menckens head spin. How can a carrier possibly distill a solution from such a bevy of complexity?

The solution is really not that difficult to understandor to implement. Each of these loss-influencing factors can be tracked and projected by class of business. Therefore, its possible to score each class of business as to the expected deterioration or improvement in its loss experience relative to the experience underlying current rates within a state.

Given currently approved rates, its then possible to rank the attractiveness of the various classes of business and to focus on those whose experience is expected to improve. In our construction example, this amounts to steering clear of construction-related workers' comp classes during period of rapid construction growth.

Construction activity and many of the factors cited above also vary significantly within larger states. Over the next few years, for example, construction in Southern California is expected to expand smartly, while it contracts in parts of Northern California. As a result, its possible to refine a workers' comp strategy below the state level to encompass variations between major metro areas.

In short, a carriers appetite can and should vary by class of business and local geographic territory.

Old H.L. probably would have been the first to point out that complex solutions to complex problems arent always correct. However, he probably would have agreed that a class- and state-specific scoring strategy would provide carriers with a useful and proactive alternative to sitting on their thumbs complaining about excessive competition.

In this article, weve looked at just some of the data “MarketStance” uses to track workers' comp conditions. In future articles, well explore some of the actual class- and territory-specific scoring techniques that are used to develop market-sensitive underwriting strategies.

Frederick Yohn is the developer of “MarketStance,” a market analysis tool for U.S. commercial property-casualty insurers and a registered trademark of IntelliStance, LLC, in Middletown, Conn.

Reproduced from National Underwriter Property & Casualty/Risk & Benefits Management Edition, August 19, 2002. Copyright 2002 by The National Underwriter Company in the serial publication. All rights reserved.Copyright in this article as an independent work may be held by the author.